Value Betting Explained: What 863 of Our Own Bets Taught Us
Value betting is the most quoted idea in football betting. We tested it on our own picks for five months and published every result. Here is the formula, a calculator, and the honest outcome.
What is value betting?
A bet has value when you believe the chance of it winning is higher than the chance the odds imply. The bookmaker’s odds can be turned into a probability: divide 1 by the decimal odds. Odds of 2.10 imply 1 / 2.10 = 47.6%. If you think the real chance is 52%, the bet has positive value on paper.
Example: 0.52 × 2.10 − 1 = +0.092, an expected return of +9.2% per bet. If your 52% is right.
That last sentence is the whole problem. The formula is simple. Knowing the true probability is not, and the rest of this page shows what happens when the estimate is off.
Value Bet Calculator
Enter the odds and your own probability to spot the edge and the expected value.
A bet has value when your probability × decimal odds > 1. We treat an edge under 5 percentage points as noise.
For education only. Odds and probabilities are your inputs; no outcome is guaranteed. 18+. Please gamble responsibly.
We tested it: 863 bets, every result public
Since May 2026 we have kept a daily table of the matches where our model disagreed most with the bookmaker market. On paper these are textbook value bets: our model’s probability times the odds, minus one, was positive for every single one. We settled all of them at the odds shown, 1 unit each, including the ones that lost.
Grouped by how much value our model claimed, the result was the opposite of what value betting promises:
| Claimed value | Bets | Won | Avg. odds | Return | |
|---|---|---|---|---|---|
| 5 to 10% | 68 | 58.8% | 1.64 | −4.6% | |
| 10 to 20% | 99 | 51.5% | 1.79 | −3.1% | |
| 20 to 40% | 346 | 41.0% | 2.19 | −7.2% | |
| 40% or more | 350 | 29.1% | 3.00 | −13.5% |
Source: our daily “largest model edge” table, 11 May to 11 October 2026, World Cup excluded. Half wins and half losses on Asian lines counted as such, voids as zero. Measured 12 October 2026.
The bigger the edge our model claimed, the more money it lost. Across all 863 bets the return was about −9%.
Why a big edge is usually a big mistake
A bookmaker’s price is not one person’s opinion. It is shaped by sharp bettors, team news and money from all over the world. When a model says the true chance is 45% and the market says 30%, there are two possibilities: the market missed something, or the model did. Our numbers say it was mostly the model.
Three things drive that:
- Missing information. Injuries, rotation, motivation and lineups move the market before they reach any dataset. A model that does not see them reports the gap as value.
- Little history. Our worst results came from early European qualifiers and smaller leagues, where the model had few past matches to learn from.
- Long odds magnify errors. The 40%+ group averaged odds of 3.00. A small overestimate of an outsider’s chance creates a huge paper edge. This is closely related to the well-known favourite-longshot bias: longshots tend to be overbet, not underpriced.
A prediction we wrote down in advance
In August, after the first three months, we noticed the losses were worst in competitions where the model had little data. We recorded a testable prediction in our internal log on 10 August 2026: once the big leagues restarted, value bets in those leagues should improve to around −4%.
| Since 15 August 2026 | Bets | Return |
|---|---|---|
| Top 7 leagues: Premier League, La Liga, Serie A, Bundesliga, Ligue 1, Eredivisie, Primeira Liga | 69 | −4.9% |
| All other competitions | 321 | −11.0% |
The prediction held for the big leagues: −4.9% against the expected −4%. But it was only half the story. Most of the table still came from other competitions, which stayed around −11%.
Better data reduced the damage. It did not create an edge.
Which markets lost least
| Market | Bets | Return | |
|---|---|---|---|
| Asian Handicap | 440 | −2.5% | |
| Double Chance | 160 | −11.1% | |
| Over/Under | 189 | −14.1% | |
| Both Teams to Score | 29 | −24.3% | |
| 1X2 (match result) | 45 | −35.8% |
Asian Handicap came closest to break-even. That fits the market: two outcomes, no draw, and quarter and half lines that soften a near miss, so there is less for a wrong estimate to lose. The 1X2 and BTTS samples are small, but both were clearly negative. Read more in our Asian Handicap guide.
What we changed
We stopped choosing picks by model value. Since 19 September 2026 our published picks follow a fixed market rule: the bookmaker’s own probability, with the margin removed, decides. Every page still shows our model’s probability next to the market’s, so you can see where they disagree, but the disagreement no longer picks the bet.
Early numbers: our main picks returned +7.8% over 26 settled bets, and the extra picks −10.7% over 18. That is far too few bets to call it proof, and we will keep publishing the record either way. You can follow it on our results page and in the tips archive.
How to use value betting sensibly
- Distrust big edges. If your number differs from the market by 10 or more percentage points, assume you missed something before you assume the bookmaker did.
- Compare with a sharp price. Bookmakers with low margins and high limits, such as Pinnacle, are the hardest prices to beat. If you consistently get better odds than their closing price (closing line value), that is a stronger sign of skill than a winning month. We wrote about our own closing line value measurement.
- Keep a full record. Every bet, at the odds you actually took. Without it you cannot tell an edge from a lucky run, and 100 bets is rarely enough.
- Prefer simple two-way markets. In our data Asian Handicap lost least and 1X2 lost most.
- Never stake what you cannot afford to lose. Even a genuine edge comes with long losing streaks. See our bankroll management guide, and if betting stops being fun, BeGambleAware offers free, confidential help.
Frequently asked questions
What is a value bet?
A bet where your estimated probability of winning is higher than the probability implied by the odds (1 divided by the decimal odds). On paper it returns a profit in the long run, but only if your estimate is more accurate than the market’s.
How do you calculate value in betting?
Multiply your probability by the decimal odds and subtract 1. With a 52% chance at odds of 2.10: 0.52 × 2.10 − 1 = +0.092, or +9.2%. Anything above zero is value on paper.
Does value betting actually work?
It can, but only with probability estimates that are better than the market. In our own test of 863 bets, the selections with the highest claimed value lost the most (−13.5%), because large gaps with the market were mostly errors in our model.
Which market is best for value betting?
In our data Asian Handicap came closest to break-even (−2.5% over 440 bets), helped by its two-way structure without a draw. 1X2 and Both Teams to Score were clearly negative.
Is value betting legal?
Yes, it is simply a way of choosing bets. Some bookmakers do limit accounts that win consistently. Only bet with operators licensed in your country and only if you are 18 or older.
Get free tips, every day
Daily AI-backed football tips in your inbox. Free, no spam, unsubscribe anytime.
18+ only. Bet responsibly. Gambling can be addictive.